Bacterial Infections in Drug Users
new england journal medicine The of review article current concepts
Bacterial Infections in Drug Users Rachel J. Gordon, M.D., M.P.H., and Franklin D. Lowy, M.D.
llicit drug use is a worldwide health problem. annually, approx- From the Division of Infectious Diseases,
1 Departments of Medicine (R.J.G., F.D.L.) imately 5 percent of the global population, or 200 million people, use illicit drugs. and Pathology (F.D.L.), Columbia Univer-
In a U.S. survey, 19.5 million people 12 years of age or older, or 8.2 percent of the sity College of Physicians and Surgeons,
i
2 population, had used an illicit drug in the prior month. Injection is one of the most New York. Address reprint requests to Dr.
Gordon at the Division of Infectious Dis- harmful routes of administration. There are an estimated 13 million injection-drug us- eases, Box 82, Columbia University Col-
3 ers worldwide, 78 percent of whom live in developing nations. lege of Physicians and Surgeons, 630 W.
4,5 Infections are among the most serious complications of drug use. Drug use plays 168th St., New York, NY 10032, or at rj216@columbia.edu. a major role in the transmission of human immunodeficiency virus (HIV), sexually transmitted diseases, and viral hepatitis. In addition to these infections, drug users risk
N Engl J Med 2005;353:1945-54. 4,6-9 Copyright © 2005 Massachusetts Medical Society. acquiring a diversity of bacterial infections. This review summarizes recent infor- mation on bacterial infections associated with drug use and examines the interactions among the drug user, the preferred drug, and the method of administration that define the nature of these infections. e p i d e m i o l o g y
Most bacterial infections among drug users are caused by the subject’s own commen-
Staphylococcus aureus
sal flora, with and streptococcus species being the most common 6-11 pathogens. Outbreaks among drug users that are caused by unusual organisms, such
Pseudomonas aeruginosa,
as clostridia species and may indicate that a particular drug or 12,13 drug-use behavior is involved.
Skin and soft-tissue infections are some of the most common infections among in- jection-drug users. Their incidence is difficult to estimate because such infections are often self-treated. A prospective study of injection-drug users in Amsterdam reported
14 an approximate incidence of one abscess per three years of injection-drug use. A cross- sectional study of injection-drug users in San Francisco found that 32 percent had an
15 abscess, cellulitis, or both, as confirmed by physical examination (Fig. 1).
Inexperience with injection may predispose the drug user to soft-tissue infection. Experienced injection-drug users who lack viable veins for use commonly resort to “skin
15 popping” (subcutaneous or intramuscular injection). Binswanger et al. reported that injection-drug users who had skin-popped within the preceding 30 days had a higher risk of soft-tissue infection than those who injected only intravenously. Injecting “speed- balls” (mixtures of cocaine and heroin), injecting more frequently, and being positive
14 for HIV infection were also associated with skin abscesses. Using dirty needles, fail- ing to clean the skin before injection, and “booting” (repeatedly flushing and pulling
16,17 back during injection) may also increase the risk of abscess.
An upsurge in skin infections, primarily abscesses, as well as more invasive infec- tions, among injection-drug users in California has been caused by community-associ-
ated methicillin-resistant (MRSA). Six of 14 patients with necrotizing fasci- itis due to community-associated MRSA at Harbor–UCLA Medical Center were current
18 S. aureus
19 or former injection-drug users. Community-associated MRSA infections have also n engl j med 353;18 www.nejm.org november 3, 2005 new england journal medicine The of
Drug users have a 10-fold increase in the risk
30 of community-acquired pneumonia. Drug us- ers, who are often smokers, may have impaired re- spiratory-clearance mechanisms as well as an in-
31 creased risk of aspiration. Immunocompromise, resulting from poor nutrition or HIV infection, may also contribute to the increased risk of respiratory
6,32,33 tract infection.
The risk of pulmonary tuberculosis (including drug-resistant cases) appears to be increased among drug users as a result of crowded living conditions (including residence in homeless shelters and drug use in “crack” cocaine houses and “shooting galler- ies”), delays in diagnosis, poor adherence to treat- ment (which increases the duration of infectious- ness), and the prevalence of HIV infection or the
34-37 acquired immunodeficiency syndrome (AIDS).
Tuberculosis and other respiratory pathogens may also be transmitted through a practice known as “shotgunning” (smoking and inhaling a drug and
Figure 1. Soft-Tissue Abscess on the Leg of an Injection-
then expelling the smoke into another person’s Drug User. mouth), a common practice among smokers of “Track” marks from injections extend along the calf.
38 crack cocaine. Inhaling cocaine also predisposes Provided courtesy of Robert Gwizdala. drug users to upper respiratory tract infections, in- cluding sinusitis and, in rare instances, septal ab-
39,40 scesses. been reported among prisoners, military recruits, A unique association exists between clostridial and men who have sex with men who use crystal infections and the use of black-tar heroin, a form of
20-24 methamphetamine. These infections, most heroin produced in Mexico. An epidemic of wound of which are positive for the mobile genetic ele- botulism occurred in California during the 1990s ment staphylococcal chromosomal cassette (SCC) in conjunction with the increased use of black-tar
mec type IV, in part reflect the clonal dissemination heroin. Most patients had injected the drug subcu-
of toxin-producing (i.e., Panton–Valentine leuko- taneously, suggesting the combined role of black- cidin) strains with a predilection to cause cutane- tar heroin and skin popping in heightening the
25
13 ous and respiratory tract infections. risk of wound botulism.
Infective endocarditis has an estimated inci- The number of tetanus cases climbed in con- dence of 1.5 to 3.3 cases per 1000 injection-drug cert with California’s epidemic of wound botu- 26,27 users per year. Infection at other sites, coloni- lism. Twenty-seven (40 percent) of the identified zation with S. aureus, and a history of infective en- cases occurred in injection-drug users, all of whom
14,26,28
41 docarditis increase the drug user’s risk. The reported subcutaneous injection of heroin. Na- risk appears augmented when the injected drug is tionally, injection-drug users accounted for 19 of cocaine, perhaps owing to the vasoconstrictive ef- the 130 cases reported from 1998 to 2000 (15 per-
42 fects of the drug or the increased frequency of in- cent). England had an outbreak of eight cases
29 jection. among injection-drug users in 2003. The close clus-
HIV infection also predisposes the drug user tering of cases suggested contamination of drugs
43 to infective endocarditis. In a Baltimore cohort of with Clostridium tetani .
2946 injection-drug users, the incidence was 3.3 cas- Necrotizing fasciitis with toxic shock syndrome es per 1000 person-years among HIV-negative sub- complicated
C. sordellii infection in at least six us-
44 jects and 13.8 cases per 1000 person-years among ers of black-tar heroin in California. Eighty-eight HIV-positive subjects. Risk factors for infective en- injection-drug users in England, Scotland, and Ire- docarditis included immunosuppression (defined land were hospitalized with a soft-tissue infection
45 as fewer than 200 CD4 cells per cubic millimeter) at an injection site and more than 30 died in 2000.
26 and more frequent injection (at least daily). Twenty-six percent of patients (9 of 35 patients who n engl j med 353;18 www.nejm.org november 3 , 2005
47 These
16 Injection at
bacterial transmission through sharing of drug paraphernalia
14,17,55 Skin popping may increase the odds of wound botulism by a factor of more than 15. Investigators have also demonstrated a dose–response relation- ship between the quantities of black-tar heroin in- troduced by skin popping and the risk of wound botulism. Cleaning the skin, cleaning old needles, p a t h o g e n e s i s
42,54 The use of speed- balls may potentiate the formation of abscesses and the growth of anaerobic organisms by caus- ing local tissue damage.
13,54 Intravenous use of black-tar heroin causes ve- nous sclerosis and promotes the practice of skin popping with the loss of usable veins. Repeated in- jection of the drug into soft tissue leads to local tissue ischemia and necrosis, establishing an anaer- obic environment that favors clostridial germina- tion and elaboration of toxin.
The drug of choice and techniques of drug prepa- ration may affect the risk of infection with particu- lar organisms. The association between black-tar heroin and clostridial infections is a case in point. Investigators believe that black-tar heroin becomes contaminated with spores when mixed with adul- terants (e.g., methamphetamine or strychnine) or diluted (“cut”) with substances such as dextrose or dyed paper. Although black-tar heroin is typically heated in water before use, clostridial spores survive boiling and may even begin to germinate.
transmission through drugs or drug
adulterants
53 These studies sug- gest that S. aureus is transmitted through drug-use networks and shared drug paraphernalia.
53 re- ported that 14 of 54 subjects who used inhalational drugs (56 percent of whom also injected drugs) had nasal colonization with S. aureus (26 percent) and that 4 of 48 pieces of their inhalational parapherna- lia were positive for S. aureus (8 percent). Closely re- lated clusters of S. aureus strains were present among users of four drug-use networks, all of whom fre- quented the same crack house.
52 Quagliarello et al.
51 This phenomenon has also been reported with Streptococcus pyogenes .
50 In Zurich, 31 injection-drug users were infected and hospitalized with the same MRSA clone.
49 In Boston, seven injection-drug users with bac- teremia caused by a single strain of S. aureus had been injected by a “street doctor” at the same “shooting gallery,” using the gallery’s drug para- phernalia.
31 In the first U.S. outbreak of community-associated MRSA among injection-drug users, the epidemic strain was cultured from a needle used by one of the pa- tients.
Drug users may also transmit bacteria by sharing drug paraphernalia. Such equipment may be rinsed only with water, or if running water is unavailable, saliva or even toilet water may be used.
n engl j med 353;18 www.nejm.org november 3, 2005
current concepts
15,31 Drug users often crush cap- sules and tablets in their mouth when preparing them for intravenous injection. They also blow out needle clots or lick their needles to facilitate injec- tion. Needle licking may double the risk of celluli- tis or abscess with oral streptococcal and anaero- bic species.
heavily colonized sites, such as the femoral vein, may also increase the risk of infection with gram- negative flora.
6,15-17 In a study of 1057 injec- tion-drug users in Baltimore, abscesses were twice as common among those who reported never clean- ing the skin as among those who reported always cleaning the skin before injection.
Poor hygiene may exacerbate the risk of infec- tion with commensal flora. Before “shooting up,” drug users often fail to clean injection sites, or do so insufficiently.
48 dem- onstrated that 12 injection-drug users hospital- ized with S. aureus infective endocarditis were in- fected with their own colonizing strain.
infections occur when commensal flora is intro- duced into the surrounding tissues or bloodstream through injection. Tuazon and Sheagren
47 perhaps because their nasal epithelium has been damaged by drug inhalation or their skin has been damaged by drug injection. Colonization with S. aureus is a risk factor for infection.
7,10,11,26,28,46 Drug users may have a higher rate of nasal or skin coloniza- tion with S. aureus than do those who do not use drugs,
The majority of nonpulmonary bacterial infec- tions among drug users are caused by S. aureus or streptococcus species.
the role of commensal flora
The bacteria responsible for infections in drug us- ers are acquired either from the drug user’s com- mensal flora or from organisms contaminating the drugs, drug adulterants, or paraphernalia. Exam- ples of these different modes of transmission and the pathogenesis of these infections are discussed below.
C. per- fringens .
C. novyi and
met the case definition) had microbiologic evidence of clostridial infections, including
15 new england journal medicine The of
or using new needles before injection were not pro- rinse syringes, combined with the solubility of tective, another indication that black-tar heroin was TaBs, was believed to explain the association be-
54
12 the source of the infection. tween TaBs and P. aeruginosa infection. A series of groin abscesses due to S. milleri was identified in
drug preparation
Scotland, where it was common to crush buprenor- Unique features of drug preparation may predis- phine (an opioid) and temazepam (a benzodiaz- pose drug users to unusual infections. The intra- epine) between the teeth and then inject it after
56 venous combination of pentazocine (an analgesic) cleaning the skin with saliva. and tripelennamine (an antihistamine), referred to as “Ts and Blues,” or TaBs, which was popular in h o s t s u s c e p t i b i l i t y
Chicago in the late 1970s, was associated with an outbreak of infective endocarditis caused by P. aeru- In addition to the role of hygiene, living condi-
12
ginosa . The tablets were crushed and then mixed tions, and tissue trauma, other factors such as mal-
with warm tap water before injection. Unlike hero- nutrition and the presence of coexisting conditions in-containing mixtures, the mixture was soluble and may impair host defenses and exacerbate the risk did not require heating before injection. The pres- of infection among drug users. HIV infection, a ence of P. aeruginosa in the tap water often used to common coexisting medical condition, increases
Table 1. Organisms Responsible for Bacterial Infections in Drug Users.* Skin, soft-tissue, and skeletal infections Staphylococcus aureus (including community-associated MRSA) a Streptococcus species — groups A, C, and G; Streptococcus anginosus ( milleri )†; and -hemolytic streptococci† Pseudomonas aeruginosa Other gram-negative bacteria ( Escherichia coli , enterobacter, klebsiella, proteus, serratia) Oral anaerobes (bacteroides species, Eikenella corrodens, fusobacterium species, peptostreptococcus species)† Mycobacterium tuberculosis Infective endocarditis‡ S. aureus (including community-associated MRSA) Streptococcus species (groups A, B, G, and others) P. aeruginosa and other gram-negative bacteria Toxin-mediated disease Clostridium botulinum, C. tetani Other clostridia species (
C. sordellii, C. novyi, C. perfringens ) Group A streptococcus and S. aureus Pulmonary infection Community-acquired pneumonia
S. pneumoniae, S. aureus, Haemophilus influenzae, Klebsiella pneumoniae, Chlamydia pneumoniae, Mycoplasma pneu- moniae, Legionella pneumophila Oropharyngeal flora (i.e., due to aspiration) Opportunistic pulmonary infections (associated with HIV disease)
M. tuberculosis (including multidrug-resistant tuberculosis), M. avium complex, P. aeruginosa, nocardia species, Rhodococcus equi Sexually transmitted infections Chlamydia trachomatis, Neisseria gonorrhoeae, Treponema pallidum, and others
- This is a selection of likely bacterial pathogens. Other bacterial or nonbacterial pathogens may also be present and should be considered in the differential diagnosis. MRSA denotes methicillin-resistant S. aureus.
† These organisms are more likely to be pathogens when oral contamination is present. ‡ Injection-drug users have an increased risk of polymicrobial infective endocarditis.
n engl j med 353;18 www.nejm.org november 3 , 2005
59 The clinical presentation of bacterial infections in
31,63 Such musculoskeletal infections are also more like- ly to be polymicrobial or anaerobic, especially if the injection-drug user contaminates the injection site, equipment, or drugs with saliva.
Culture of a biopsy specimen of the involved bone grew Enterobacter cloacae .
Figure 2. Sagittal View of Magnetic Resonance Image in an Injection-Drug User with Vertebral Osteomyelitis Involving T6–T7 (Arrow) with an Adjacent Epidural Abscess.
4,6,30,33 A thorough medical history taking includes questions regarding tuberculosis, HIV, and risk factors for aspiration. Aspiration pneu- monia as well as pneumonia caused by S. pneumoni- c l i n i c a l f e a t u r e s
Respiratory tract infections are among the most frequent sequelae of drug use, and patients may present with atypical clinical and radiographic find- ings.
respiratory tract infections
28,64,65
64 If infective en- docarditis is left-sided, complications such as brain and splenic abscess may occur owing to septic em- bolization. Infective endocarditis associated with injection-drug use is also one of the rare settings in which polymicrobial endocarditis is encoun- tered.
9,27,28 HIV-infected injection- drug users with fewer than 200 CD4 cells per cubic millimeter have a higher mortality rate than those with less advanced HIV disease.
Endovascular infections, including infective en- docarditis, septic thrombophlebitis, mycotic an- eurysms, and sepsis, are among the most serious complications of injection-drug use. An endovas- cular source should be investigated in all injec- tion-drug users with bacteremia. Although left-sid- ed infective endocarditis in drug users is similar to that seen in non–drug users, infective endocardi- tis in injection-drug users is most commonly due to S. aureus and involves the tricuspid valve in 70 percent of patients. Symptoms may include fever, dyspnea, pleuritic chest pain, and cough. A murmur may be absent. The mortality rate in these cases is less than 5 percent.
endovascular infections
n engl j med 353;18 www.nejm.org november 3, 2005
current concepts drug users’ susceptibility to bacterial infections.
62,63 These infections may result, in part, from injecting in high-risk areas such as the jugular vein
62,63 In- jection-drug users are susceptible to skeletal infec- tions in uncommon places such as the sternocla- vicular and sacroiliac joints, in addition to more usual sites, such as the vertebral spine (Fig. 2) and knee.
Musculoskeletal infections, including septic ar- thritis and osteomyelitis, generally result from he- matogenous seeding or, less commonly, local ex- tension of a skin or soft-tissue infection in drug users. These infections may be indolent, and the only symptom may be pain without fever.
musculoskeletal infections
60,61
21,23 Community-associated MRSA should be considered as a potential cause in patients who have a history of such infections, patients in potential risk groups who live in high-risk regions who have abscesses (especially complicated ones), and patients with infections that are not respond- ing to conventional antibiotics (photographs of ab- scesses are available at http://lapublichealth.org/ acd/docs/mrsa_colorenhanced.pdf ).
Although the abscesses associated with communi- ty-acquired MRSA generally appear similar to other abscesses, they have occasionally been confused with spider bites.
skin and soft-tissue infections
drug users is generally similar to that encountered in patients who do not use drugs. This section will highlight features that are unique to the drug user. Bacteria associated with infections in drug users are noted in Table 1.
58 others have also noted an increased risk of death from bacterial infection among HIV-infected drug users.
36,57 First reported by Stoneburner et al.,
(“pocket shot”) and femoral vein (“groin hit”). n engl j med 353;18 www.nejm.org november 3, 2005
The new england journal of medicine ae, Haemophilus influenzae, S. aureus, and Klebsiella pneu- moniae are among the most common reasons for
S. aureus : dicloxacillin or cephalexin,‡ 500 mg every 6 hr
For osteomyelitis, serious infections, and possible MRSA infection, add vancomycin‡ or teico- planin‡¶ Acute right-sided infective endocarditis Diagnosis is based on the modified Duke criteria. 71 Culture of multiple blood specimens before the initiation of antibiotic therapy is the optimal approach.
Ampicillin–sulbactam,‡ 1.5–3.0 g every 6 hr, plus gentamicin,‡ 1.5–2.0 mg/kg every 8 hr for seri- ous or complicated in- fections; piperacillin– tazobactam,‡ 3.375 g ev- ery 4–6 hr or 4.5 g every 6–8 hr; ticarcillin–clavu- lanate,‡ 3.1 g every 4–6 hr; cefepime,‡ 1–2 g every 12 hr;
Incision and drainage when appropriate; wound care Amoxicillin–clavulanate,‡ 875 mg every 12 hr; For serious penicillin allergy: clindamycin and quino- lone (dose and route based on type and sever- ity of infection)
Send specimens for Gram’s staining, culture, and susceptibility testing. Consider imaging to di- agnose or define deep- seated infections. A bone biopsy is important when osteomyelitis is suspected regardless of whether blood-culture results are positive. Specimens for anaero- bic culture require spe- cial handling.
If MRSA suspected: vancomycin,‡ 15 mg/kg every 12 hr; teicopla- nin,‡¶ 6 mg/kg every 12 hr for 3 doses, then 6 mg/kg every 24 hr; linezolid, 600 mg every 12 hr; daptomycin,‡ 4 mg/kg every 24 hr Infections in which oral con- tamination is suspected, including skin or soft-tis- sue and skeletal infec- tions(septic arthritis and bursitis, tenosynovitis, and osteomyelitis)
TMP-SMX,‡ 6–10 mg/kg of body weight/day (TMP) in divided doses given every 8–12 hr; clin- damycin,§ 300 mg every 6 hr or 450 mg every 8 hr; doxycycline or minocy- cline, 100 mg every 12 hr; linezolid, 600 mg every 12 hr
For methicillin-susceptible S. aureus : nafcillin or ox- acillin, 1–2 g every 4–6 hr; cefazolin,‡ 1–2 g ev- ery 8 hr If MRSA suspected:
Incision and drainage plus wound care may suffice for un- complicated abscesses For methicillin-susceptible
hospitalization.
Send drainage for Gram’s staining, culture, and susceptibility testing.
Table 2. Initial Management of Bacterial Infectious Syndromes among Suspected Drug Users.* Clinical Scenario Selected Diagnostic Tests Empirical Treatment Options† Oral Parenteral Skin or soft-tissue infection in which S. aureus is a likely pathogen
Among the most important of the unusual infec- tions related to drug use are the clostridial infec- tions that have been associated with the use of black- tar heroin. These may have typical presentations, as with the cases of tetanus, or an unusual presenta-
miscellaneous infections
66
7,31 Tuberculosis must be considered in drug users presenting with pneumonia and known or suspect- ed HIV infection. Drug users who have tuberculosis associated with HIV disease can present atypi- cally (i.e., without cavitary lesions), and the puri- fied-protein-derivative test may be negative. Hilar or mediastinal lymphadenopathy may be the only finding.
4,5,32,33 Undiagnosed HIV can first become evident as the result of an opportunistic in- fection such as pneumocystis pneumonia. Endovas- cular infections can cause septic pulmonary emboli producing pneumonias and lung abscesses.
Vancomycin,‡ 15 mg/kg IV every 12 hr (or teicoplanin,‡¶ 12 mg/kg every 12 hr for 3 doses, then 12 mg/kg every 24 hr), plus gentamicin,‡ 1 mg/kg every 8 hr or consid- er nafcillin or oxacillin, 2 g IV every 4 hr, plus gentami- cin,‡ 1 mg/kg every 8 hr, if MRSA not present in the community; consider broadening coverage (pseudomonal, gram-negative, or fungal antibiotics) on the basis of patient risk factors¿ current concepts Table 2. (Continued.) Clinical Scenario Selected Diagnostic Tests Empirical Treatment Options†
Oral Parenteral Pulmonary infection (com- Radiographic imaging: Hospitalized with community-acquired pneumonia: ceftri- munity-acquired pneu- Gram’s and AFB staining axone, 1–2 g IV every 24 hr, and azithromycin, 500 mg monia and aspiration of sputum and cultures
IV every 24 hr or respiratory fluoroquinolone** pneumonia, pulmonary of sputum and blood. Aspiration pneumonia likely: clindamycin, 600–900 mg IV tuberculosis, and other Bronchoscopy may be
every 8 hr
opportunistic patho- needed to diagnosePneumocystis pneumonia suspected: TMP–SMX,‡ 15–20 gens) in drug users, in- pneumocystis pneu- mg/kg/day (TMP dose), given in divided doses every cluding those with HIV monia. 6–8 hr (with or without corticosteroids) or AIDS or risk factors In certain cases, performing
For tuberculosis, see treatment guidelines at for HIV infection a PPD test or checking www.thoracic.org or www.who.int/tb/en/index.html for Streptococcus pneu- moniae and legionella urinary antigens may be useful.
Presentation involving sep- Gram’s staining, culture, For botulism: trivalent antitoxin (type A, B, or E), 1 vial, tic or neurologic findings and susceptibility testing available from the appropriate public health authority; of unknown cause with should be done if appli- and penicillin G,‡ 3 million U IV every 6 hr or without skin or soft- cable.
For tetanus-prone wounds or tetanus: human tetanus im- tissue infection mune globulin and tetanus toxoid; metronidazole, 500 mg orally or IV every 8 hr
For other clostridia species (may be polymicrobial): débridement of skin or soft-tissue infections; ampicil- lin–sulbactam,‡ 3 g IV every 6 hr plus vancomycin,‡ 15 mg/kg IV every 12 hr (or teicoplanin‡¶)††
Sexually transmitted infec- Examination and workup are Follow CDC treatment guidelines or those of public health tions conducted according to authorities (available at www.who.int/topics/ local health department sexually_transmitted_infections/en/) guidelines. RPR and
VDRL tests may be false positive; confirm results with FTA-ABS test.
- MRSA denotes methicillin-resistant S. aureus, TMP-SMX trimethoprim–sulfamethoxazole, AFB acid-fast bacilli, RPR rapid plasma reagin, VDRL Venereal Disease Research Laboratory, FTA-ABS fluorescent treponemal antibody absorption, IV intravenously, and CDC Centers for Disease Control and Prevention.
† Therapy should be adjusted on the basis of the culture results and antibiotic susceptibilities. ‡ The dose must be adjusted in patients with reduced creatinine clearance. § Clindamycin should not be used if the isolate is resistant to erythromycin. ¶ This drug is not available in the United States.
72
¿ Therapy is generally continued for four to six weeks. Baddour et al. provide specific recommendations.- Respiratory quinolones include gatifloxacin, levofloxacin, and moxifloxacin. ††Hyperbaric oxygen has also been used.
tion, as with the cases of C. sordellii infection in which be possible. Numerous risk-reducing strategies may patients presented with toxic shock syndrome (with help prevent bacterial infections among drug us- 44,45 or without cutaneous lesions). Patients should ers, particularly among new users, the ones at great- be asked about the use of subcutaneous and intra- est risk. These programs include the provision of muscular injections, the use of black-tar heroin, medically supervised injection facilities, syringe- and the date of the last tetanus immunization. Seri- exchange programs, and street-based education pro- ous toxin-mediated disease, such as botulism, may grams directed at the use of sterile injection prac-
67,68 masquerade as intoxication (e.g., slurred speech). tices. Safer injection practices include boiling the drug, cleaning the skin with alcohol and the drug paraphernalia with bleach, avoiding contam- p r e v e n t i o n inating or sharing needles with others, and avoid-
Eliminating drug use is the surest way to control ing the use of dangerous injection sites such as the 17,69 associated infections, but this goal may not always neck and groin. High-risk behavior such as hav- n engl j med 353;18 www.nejm.org november 3, 2005 n engl j med 353;18 www.nejm.org november 3, 2005
The new england journal of medicine
7. Cherubin CE, Sapira JD. The medical
York: United Nations Office on Drugs and Crime, 2004.
4. Scheidegger C, Zimmerli W. Infectious
complications in drug addicts: seven-year review of 269 hospitalized narcotics abusers in Switzerland. Rev Infect Dis 1989;11:486- 93. [Erratum, Rev Infect Dis 1990;12:165.]
5. Palepu A, Tyndall MW, Leon H, et al. Hos-
pital utilization and costs in a cohort of in- jection drug users. CMAJ 2001;165:415-20.
6. Louria DB, Hensle T, Rose J. The major medical complications of heroin addiction.
Ann Intern Med 1967;67:1-22.
complications of drug addiction and the medical assessment of the intravenous drug user: 25 years later. Ann Intern Med 1993; 119:1017-28.
Health. Rockville, Md.: Office of Applied Studies, Substance Abuse and Mental Health Services Administration, 2004.
8. Levine DP, Sobel JD, eds. Infections in
intravenous drug abusers. New York: Ox- ford University Press, 1991.
9. Levine DP, Brown PV. Infections in in-
jection drug users. In: Mandell GL, Bennett JE, Dolin R, eds. Mandell, Douglas, and Bennett’s principles and practice of infec- tious diseases. 6th ed. Vol. 2. Philadelphia: Churchill Livingstone, 2005:3462-76.
10. Crane LR, Levine DP, Zervos MJ, Cum-
mings G. Bacteremia in narcotic addicts at the Detroit Medical Center. I. Microbiology, epidemiology, risk factors, and empiric ther- apy. Rev Infect Dis 1986;8:364-73.
11. Bassetti S, Battegay M. Staphylococcus au- reus infections in injection drug users: risk
factors and prevention strategies. Infection 2004;32:163-9.
3. World drug report. Vol. 1. Analysis. New
2. 2003 National Survey on Drug Use and
ing unprotected sex and sex with multiple partners, which facilitates the transmission of HIV, sexually transmitted diseases, and community-associated MRSA, should be discouraged.
77 The choice of therapy also requires consider- ation of regional differences in antibiotic suscep- tibility among prevalent bacteria. With the emer- gence of community-associated MRSA as a virulent pathogen, it is crucial to obtain cultures for sensi- tivity testing before therapy is initiated.
22 In a Baltimore study of injection-drug users, 60 percent reported a history of a sexually transmitted infection.
70 Vac- cination with pneumococcal, H. influenzae type b, and tetanus vaccines
41,42 as well as routine screen- ing for tuberculosis, HIV infection, and sexually transmitted diseases may help prevent bacterial in- fection and disease transmission.
The medical management of bacterial infections in drug users begins with the recognition of drug use and its associated coexisting conditions. The spe- cific issues of the management of drug withdraw- al, adherence to therapy, and difficulties of intra- venous access must be a part of the therapeutic strategy. Close attention to local outbreaks and bac- terial antibiotic-resistance profiles is important. Management suggestions for selected syndromes related to drug use are included in Table 2.
Although conventional treatment is suggested, short courses of parenteral and oral regimens have been investigated for right-sided infective endo- carditis associated with injection-drug use. Selected cases of right-sided infective endocarditis due to methicillin-susceptible S. aureus have been success- fully treated with a penicillinase-resistant penicil- lin and an aminoglycoside (for two weeks) or cipro- floxacin and rifampin (for four weeks). Exclusion criteria for short-course or oral therapy for right- sided S. aureus infective endocarditis have included infections caused by MRSA, complicated cases char- acterized by a slow clinical response (more than 96 hours), large valvular vegetations (more than 2 cm), extracardiac complications, or injection-drug use in a patient with AIDS.
27,73-75 For regimens that re- quire prolonged therapy such as those for tubercu- losis, the use of directly observed therapy is now standard.
76 For drug users with sexually transmit- ted diseases, in addition to the use of standard rec- ommendations, the use of single-dose therapy when possible has been successful.
18,60 As we noted earlier, unlikely pathogens must be consid- ered on the basis of the patient’s drug-use behav- ior, including the type of drug used and the route of administration.
York: United Nations Office on Drugs and Crime, 2005.
Injection-drug users who report having a fever with no localizing findings pose a particular diag- nostic problem. Infection is just one of many po- tential diagnoses in these patients. Infective en- docarditis will be diagnosed in up to 20 percent of injection-drug users hospitalized with fever.
29 However, the diagnosis of infective endocarditis in these patients is difficult when based solely on clinical and laboratory criteria.
78 Therefore, it has been recommended that these patients be admit- ted for further evaluation. Hospitalization of drug users also provides an important opportunity to of- fer comprehensive medical care.
Supported by grants (DA-09655 and DA-15018) from the Nation- al Institute on Drug Abuse and by a grant (CCR223380-01) from the Centers for Disease Control and Prevention.
Dr. Gordon reports having received grant support from Glaxo- SmithKline. Dr. Lowy reports having received consulting or lecture fees from Wyeth, Pfizer, and Cubist; having received grant support from GlaxoSmithKline; and holding stock in Merck and Pfizer.
We are indebted to Christine Kubin, Michael Tapper, and Abigail Zuger for their critical review of this manuscript, and to Robert Gwizdala for providing Figure 1. t r e a t m e n t
r e f e r e n c e s
1. World drug report. Vol. 1. Analysis. New
12. Shekar R, Rice TW, Zierdt CH, Kallick n engl j med 353;18 www.nejm.org november 3, 2005
current concepts
40. Silverman HS, Smith AL. Staphylococ- cal sepsis precipitated by cocaine sniffing.
G, Vargas Y, Vugia DJ. Outbreak of necrotiz- ing fasciitis due to Clostridium sordelli among black-tar heroin users. Clin Infect Dis 2004; 38:e87-e91.
44. Kimura AC, Higa JI, Levin RM, Simpson
drug users in England: update. Eurosurvell Wkly 2003;7:5-7.
43. Cluster of cases of tetanus in injecting
Cortese MM, Murphy TV. Tetanus surveil- lance — United States, 1998–2000. MMWR Surveill Summ 2003;52:1-8.
42. Pascual FB, McGinley EL, Zarardi LR,
California, 1997. MMWR Morb Mortal Wkly Rep 1998;47:149-51.
41. Tetanus among injecting-drug users —
N Engl J Med 1985;312:1706.
complications of cocaine abuse. In: Galant- er M, ed. Recent developments in alcohol- ism: alcohol and cocaine. Vol. 10. New York: Plenum Press, 1992:363-77.
plained illness and death among injecting- drug users — Scotland, Ireland, and En- gland. MMWR Morb Mortal Wkly Rep 2000; 49:543-5.
39. Perper JA, Van Thiel DH. Respiratory
Paone D, Salomon N, Des Jarlais DC. Doing a shotgun: a drug use practice and its rela- tionship to sexual behaviors and infection risk. Soc Sci Med 1999;48:1441-8.
38. Perlman DC, Henman AR, Kochems L,
et al. Unsuspected recent transmission of tuberculosis among high-risk groups: im- plications of universal tuberculosis geno- typing in its detection. Clin Infect Dis 2005; 40:366-73.
37. Malakmadze N, Gonzalez IM, Oemig T,
use and human immunodeficiency virus in- fection. Clin Infect Dis 1995;20:1467-79.
36. Alcabes P, Friedland G. Injection drug
Spread of Mycobacterium tuberculosis in a com- munity implementing recommended ele- ments of tuberculosis control. JAMA 2000; 283:2968-74.
35. Chin DP, Crane CM, Diul MY, et al.
tuberculosis — Contra Costa County, Cali- fornia, 1987–1990. MMWR Morb Mortal Wkly Rep 1991;40:485-9.
45. Update: Clostridium novyi and unex-
46. Summanen PH, Talan DA, Strong C, et
K, et al. Bacterial pneumonia in HIV-infect- ed patients: analysis of risk factors and prognostic indicators. J Acquir Immune Def- ic Syndr Hum Retrovirol 1998;18:39-45.
bacteremia linked to intrave- nous drug abusers using a “shooting gal- lery.” Am J Med 1986;80:770-6.
multifocal skin necrosis: synergism between invasive streptococcal infection and cocaine- induced tissue ischaemia? Acta Derm Ve- nereol 1996;76:239-41.
55. Hoeger PH, Haupt G, Hoelzle E. Acute
MacKenzie WR, Vugia DJ. Wound botulism associated with black tar heroin among in- jecting drug users. JAMA 1998;279:859- 63.
54. Passaro DJ, Werner SB, McGee J,
al. Strains of Staphylococcus aureus obtained from drug-use networks are closely linked. Clin Infect Dis 2002;35:671-7.
53. Quagliarello B, Cespedes C, Miller M, et
Aebi C, Tauber MG. Outbreak among drug users caused by a clonal strain of group A streptococcus. Emerg Infect Dis 2000;6: 175-9.
52. Bohlen LM, Muhlemann K, Dubuis O,
Epidemic spread of a single clone of methi- cillin-resistant Staphylococcus aureus among injection drug users in Zurich, Switzerland. Clin Infect Dis 2001;32:581-6.
51. Fleisch F, Zbinden R, Vanoli C, Ruef C.
coccus aureus
al. Bacteriology of skin and soft-tissue infec- tions: comparison of infections in intrave- nous drug users and individuals with no his- tory of intravenous drug use. Clin Infect Dis 1995;20:Suppl 2:S279-S282.
McCabe WR. Methicillin-resistant Staphylo-
50. Craven DE, Rixinger AI, Goularte TA,
epidemiologic observa- tions during a community-acquired out- break. Ann Intern Med 1982;96:11-6.
Staphylococcus aureus:
Pohlod D, Fisher E. Methicillin-resistant
49. Saravolatz LD, Markowitz N, Arking L,
cal endocarditis in parenteral drug abusers: source of the organism. Ann Intern Med 1975;82:788-90.
48. Tuazon CU, Sheagren JN. Staphylococ-
H. Nasal carriage of Staphylococcus aureus: ep- idemiology, underlying mechanisms, and associated risks. Clin Microbiol Rev 1997; 10:505-20.
47. Kluytmans J, van Belkum A, Verbrugh
34. Crack cocaine use among persons with
33. Tumbarello M, Tacconelli E, de Gaetano
CA. Outbreak of endocarditis caused by
factors for skin and soft-tissue abscesses among injection drug users: a case-control study. Clin Infect Dis 2001;33:35-40.
21. Public health dispatch: outbreaks of
skin or soft tissue infections in a state prison — Mississippi, 2000. MMWR Morb Mortal Wkly Rep 2001;50:919-22.
20. Methicillin-resistant Staphylococcus au- reus
in Los Angeles. N Engl J Med 2005;352:1445-53.
ylococcus aureus
G, et al. Necrotizing fasciitis caused by com- munity-associated methicillin-resistant Staph-
19. Miller LG, Perdreau-Remington F, Rieg
drug users — San Francisco, California, 1996–2000. MMWR Morb Mortal Wkly Rep 2001;50:381-4.
18. Soft tissue infections among injection
17. Murphy EL, DeVita D, Liu H, et al. Risk
Staphylococcus aureus
Nelson KE. Bacterial infections and skin cleaning prior to injection among intrave- nous drug users. Public Health Rep 1992; 107:595-8.
16. Vlahov D, Sullivan M, Astemborski J,
RN, Rybold DJ, Edlin BR. High prevalence of abscesses and cellulitis among community- recruited injection drug users in San Fran- cisco. Clin Infect Dis 2000;30:579-81.
15. Binswanger IA, Kral AH, Bluthenthal
GH, Coutinho RA, van den Hoek A. Human immunodeficiency virus infection and other risk factors for skin abscesses and endocar- ditis among injection drug users. J Clin Epi- demiol 1996;49:1149-54.
14. Spijkerman IJ, van Ameijden EJ, Mientjes
Schechter R, Vugia DJ. Wound botulism in California, 1951-1998: recent epidemic in heroin injectors. Clin Infect Dis 2000;31: 1018-24.
13. Werner SB, Passaro D, McGee J,
serotype O11 among pentazocine and tripelennamine abusers in Chicago. J Infect Dis 1985;151:203-8.
Pseudomonas aeruginosa
community-associated methicillin-resistant
skin infections — Los Angeles County, California, 2002–2003. MMWR Morb Mortal Wkly Rep 2003;52:88.
al. Community-acquired pneumonia in a co- hort of former injection drug users with and without human immunodeficiency virus in- fection: incidence, etiologies, and clinical aspects. Clin Infect Dis 1996;23:107-13.
27. Miro JM, del Rio A, Mestres CA. Infec-
32. Boschini A, Smacchia C, Di Fine M, et
travenous drug use. J Gen Intern Med 1990; 5:249-57.
31. Stein MD. Medical complications of in-
intravenous drug misuse. 2. Infective and HIV related complications. Thorax 1990;45: 957-61.
30. Hind CR. Pulmonary complications of
Modin G. Cocaine use and the risk for endo- carditis in intravenous drug users. Ann In- tern Med 1987;106:833-6.
29. Chambers HF, Morris DL, Tauber MG,
carditis in intravenous drug users: prognos- tic features in 102 episodes. Ann Intern Med 1992;117:560-6.
28. Hecht SR, Berger M. Right-sided endo-
tive endocarditis in intravenous drug abus- ers and HIV-1 infected patients. Infect Dis Clin North Am 2002;16:273-95.
Freedman TL, Vlahov D. Prospective study of infective endocarditis among injection drug users. J Infect Dis 2002;185:1761-6.
22. Ly LT, Revuelta MP, Hongo I, Kreiswirth
26. Wilson LE, Thomas DL, Astemborski J,
al. Involvement of Panton-Valentine leuko- cidin-producing Staphylococcus aureus in pri- mary skin infections and pneumonia. Clin Infect Dis 1999;29:1128-32.
25. Lina G, Piemont Y, Godail-Gamot F, et
Emerg Infect Dis 2004;10:941-4.
Staphylococcus aureus among military recruits.
Community-acquired methicillin-resistant
24. Zinderman CE, Conner B, Malakooti MA, LaMar JE, Armstrong A, Bohnker BK.
infections in correctional facilities — Georgia, California, and Texas, 2001–2003. MMWR Morb Mortal Wkly Rep 2003;52: 992-6.
23. Methicillin-resistant Staphylococcus au- reus
BN, Davis S, Saltzman BR. Clonal outbreak of community-acquired methicillin resis- tant Staphylococcus aureus skin abscesses in men who have sex with men in New York City: possible association with crystal meth- amphetamine usage. Presented at the annu- al meeting of the Infectious Diseases Society of America, Boston, September 30–October 3, 2004. abstract.
56. Hemingway DM, Balfour AE, McCart- n engl j med 353;18 www.nejm.org november 3, 2005
current concepts
coccus aureus
providers regarding advice to persons who inject illicit drugs. MMWR Morb Mortal Wkly Rep 1997;46:510.
70. Nelson KE, Vlahov D, Cohn S, et al. Sex-